Early diagnosis and efficient therapy is essential for the prevention of micro and macrovascular complications of diabetes mellitus (DM). Different diagnostic criteria for DM have been proposed by different international organizations. In this study we compared criteia of "National Diabetes Data Group" (NDDG,1979), "American Diabetes Association" (ADA ,1997) and "World Health Organisation" (WHO, 1999) by retrospectively revising our 840 patients (520 women, 320 men) who had undergone oral glucose tolerance testing in the last 2 years. We determined prevalances of diabetes and impaired glucose tolerance according to those 3 groups of criteria. Prevalance of diabetes was found as %6.1 according to NDDG; as %6.9 according to WHO and %1.6 according to ADA criteria.The results of the present study demonstrated that oral glucose tolerance testing, when evaluated according to WHO criteria, can diagnose more diabetics and in an earlier stage. WHO criteria can be used in confidence instead of NDDG criteria which are also cumbersome to apply. Finally, evaluation of the OGTT data by the ADA criteria, resulted in an underestimation of diabetes.